Melissa J Palmer and colleagues recently published their review. Read the plain language statement below or see the full review on the Cochrane Library.
Background
The burden of poor sexual and reproductive health (SRH) worldwide is substantial, disproportionately affecting those living in low‐ and middle‐income countries. Targeted client communication (TCC) delivered via mobile devices (MD) (TCCMD) may improve the health behaviours and service use important for sexual and reproductive health.
Objectives
To assess the effects of TCC via MD on adolescents' knowledge, and on adolescents’ and adults' sexual and reproductive health behaviour, health service use, and health and well‐being.
Search methods
In July/August 2017, we searched five databases including The Cochrane Central Register of Controlled Trials, MEDLINE and Embase. We also searched two trial registries. A search update was carried out in July 2019 and potentially relevant studies are awaiting classification.
Selection criteria
We included randomised controlled trials of TCC via MD to improve sexual and reproductive health behaviour, health service use, and health and well‐being. Eligible comparators were standard care or no intervention, non‐digital TCC, and digital non‐targeted communication.
Data collection and analysis
We used standard methodological procedures recommended by Cochrane, although data extraction and risk of bias assessments were carried out by one person only and cross‐checked by a second. We have presented results separately for adult and adolescent populations, and for each comparison.
Main results
We included 40 trials (27 among adult populations and 13 among adolescent populations) with a total of 26,854 participants. All but one of the trials among adolescent populations were conducted in high‐income countries. Trials among adult populations were conducted in a range of high‐ to low‐income countries. Among adolescents, nine interventions were delivered solely through text messages; four interventions tested text messages in combination with another communication channel, such as emails, multimedia messaging, or voice calls; and one intervention used voice calls alone. Among adults, 20 interventions were delivered through text messages; two through a combination of text messages and voice calls; and the rest were delivered through other channels such as voice calls, multimedia messaging, interactive voice response, and instant messaging services.
Adolescent populations
TCCMD versus standard care
TCCMD may increase sexual health knowledge (risk ratio (RR) 1.45, 95% confidence interval (CI) 1.23 to 1.71; low‐certainty evidence). TCCMD may modestly increase contraception use (RR 1.19, 95% CI 1.05 to 1.35; low‐certainty evidence). The effects on condom use, antiretroviral therapy (ART) adherence, and health service use are uncertain due to very low‐certainty evidence. The effects on abortion and STI rates are unknown due to lack of studies.
TCCMD versus non‐digital TCC (e.g. pamphlets)
The effects of TCCMD on behaviour (contraception use, condom use, ART adherence), service use, health and wellbeing (abortion and STI rates) are unknown due to lack of studies for this comparison.
TCCMD versus digital non‐targeted communication
The effects on sexual health knowledge, condom and contraceptive use are uncertain due to very low‐certainty evidence. Interventions may increase health service use (attendance for STI/HIV testing, RR 1.61, 95% CI 1.08 to 2.40; low‐certainty evidence). The intervention may be beneficial for reducing STI rates (RR 0.61, 95% CI 0.28 to 1.33; low‐certainty evidence), but the confidence interval encompasses both benefit and harm. The effects on abortion rates and on ART adherence are unknown due to lack of studies.
We are uncertain whether TCCMD results in unintended consequences due to lack of evidence.
Adult populations
TCCMD versus standard care
For health behaviours, TCCMD may modestly increase contraception use at 12 months (RR 1.17, 95% CI 0.92 to 1.48) and may reduce repeat abortion (RR 0.68 95% CI 0.28 to 1.66), though the confidence interval encompasses benefit and harm (low‐certainty evidence). The effect on condom use is uncertain. No study measured the impact of this intervention on STI rates. TCCMD may modestly increase ART adherence (RR 1.13, 95% CI 0.97 to 1.32, low‐certainty evidence, and standardised mean difference 0.44, 95% CI ‐0.14 to 1.02, low‐certainty evidence). TCCMD may modestly increase health service utilisation (RR 1.17, 95% CI 1.04 to 1.31; low‐certainty evidence), but there was substantial heterogeneity (I2 = 85%), with mixed results according to type of service utilisation (i.e. attendance for STI testing; HIV treatment; voluntary male medical circumcision (VMMC); VMMC post‐operative visit; post‐abortion care). For health and well‐being outcomes, there may be little or no effect on CD4 count (mean difference 13.99, 95% CI ‐8.65 to 36.63; low‐certainty evidence) and a slight reduction in virological failure (RR 0.86, 95% CI 0.73 to 1.01; low‐certainty evidence).
TCCMD versus non‐digital TCC
No studies reported STI rates, condom use, ART adherence, abortion rates, or contraceptive use as outcomes for this comparison. TCCMD may modestly increase in service attendance overall (RR: 1.12, 95% CI 0.92‐1.35, low certainty evidence), however the confidence interval encompasses benefit and harm.
TCCMD versus digital non‐targeted communication
No studies reported STI rates, condom use, ART adherence, abortion rates, or contraceptive use as outcomes for this comparison. TCCMD may increase service utilisation overall (RR: 1.71, 95% CI 0.67‐4.38, low certainty evidence), however the confidence interval encompasses benefit and harm and there was considerable heterogeneity (I2 = 72%), with mixed results according to type of service utilisation (STI/HIV testing, and VMMC).
Few studies reported on unintended consequences. One study reported that a participant withdrew from the intervention as they felt it compromised their undisclosed HIV status.
Authors' conclusions
TCCMD may improve some outcomes but the evidence is of low certainty. The effect on most outcomes is uncertain/unknown due to very low certainty evidence or lack of evidence. High quality, adequately powered trials and cost effectiveness analyses are required to reliably ascertain the effects and relative benefits of TCC delivered by mobile devices. Given the sensitivity and stigma associated with sexual and reproductive health future studies should measure unintended consequences, such as partner violence or breaches of confidentiality.